Looking for help with the Mann-Whitney U Test

S^4

New Member
#1
Hey, I'm a highschool student finishing up with AP Statistics. In my final project, I am using the Mann-Whitney U Test to compare two un-normal data sets, but this test is beyond the scope of the course, and I have never really learned how to do it. The descriptions I have found on how to carry out this test have been incomplete or inconsistent. The majority of the information I have found has been consistent and successful, but I'm not entirely sure how to calculate the U-Value once the data from my two sets is ranked.

One source told me that for each data set, I should take each individual data point, and add up all of the data points in the other data set that are ranked higher than that point, then take sum of all of these, and do the same for both data sets. Then use the smaller as my U-Value. Using this method gave me a test statistic of -30, which seemed very unreasonable.

Thanks in advance
 

S^4

New Member
#4
Thanks, I found my mistake, I counting ties as 0 points, when they should've been .5s, and my data had A LOT of ties, which explained the ridiculously small test statistic.